Content uploaded by David Casey
Author content
All content in this area was uploaded by David Casey
Content may be subject to copyright.
Preventing Bullying and Sexual Harassment
in Elementary Schools:
The Expect Respect Model
Ellen Sanchez
Trina Reed Robertson
Carol Marie Lewis
Barri Rosenbluth
Tom Bohman
David M. Casey
SUMMARY. The Expect Respect Elementary School Project, funded
by the U. S. Centers for Disease Control and Prevention, established a
Ellen Sanchez, MEd, AASECT, is Project Coordinator; Trina Reed Robertson,
MA, is Project Evaluator, Expect Respect: Elementary School Project. Carol Marie
Lewis, PhD, is Associate Director, Center for Social Work Research, University of
Texas-Austin. Barri Rosenbluth, LMSW-ACP, is Director of School Based Services,
SafePlace: Domestic Violence and Sexual Assault Survival Center. Tom Bohman,
PhD, is Manager, Research Consulting, Academic Computing and Instructional Tech
-
nology Services, University of Texas-Austin. David M.Casey, PhD, is Human Devel
-
opment and Family Sciences, University of Texas-Austin.
Address correspondence to: Ellen Sanchez, SafePlace, P. O. Box 19454, Austin, TX
78760 (E-mail:esanchez@austin-safeplace.org).
This reportwas prepared on contract with the Family and Intimate Partner Violence
Prevention Team of the National Centers for Disease Control and Prevention. The
views expressed are those of the authors and do not necessarily represent the views of
the Centers for Disease Control and Prevention.
[Haworth co-indexing entry note]: “Preventing Bullying and Sexual Harassment in Elementary Schools:
The Expect Respect Model.” Sanchez, Ellen et al. Co-published simultaneously in Journal of Emotional
Abuse (The Haworth Maltreatment & Trauma Press, an imprint of The Haworth Press, Inc.) Vol. 2, No. 2/3,
2001, pp. 157-180; and: Bullying Behavior: Current Issues, Research, and Interventions (ed: Robert A.
Geffner, Marti Loring, and Corinna Young) The Haworth Maltreatment & Trauma Press, an imprint of The
Haworth Press, Inc., 2001, pp. 157-180. Single or multiple copies of this article are available for a fee from
The Haworth Document Delivery Service [1-800-HAWORTH, 9:00 a.m. - 5:00 p.m. (EST). E-mail address:
getinfo@haworthpressinc.com].
2001 by The Haworth Press, Inc. All rights reserved. 157
model fortheprimary prevention of datingviolenceby addressing bully
-
ing and sexual harassment on public school campuses. This three-year
projectwasbased onthe belief thatwhen bullyingandsexual harassment
go unchecked in elementary schools, these behaviors condition students
to accept mistreatment in their peer relationships, laying the foundation
for abuse in future dating relationships. The Expect Respect Project as
-
sisted six public elementary schools in Austin, TX in taking a whole
school approach to stopping bullying, sexual harassment and gender vio
-
lence by providing staff training, classroom education, parent education,
assistance with policy development, and support services. Data from the
first year of implementation showed students in the intervention schools
had asignificant increase in their:(1)ability to identifysexualharassment;
(2) awareness of school policy to protect them from sexual harassment by
other students; and, (3) willingness to intervene on behalf of another stu-
dent. Thisarticle represents the project’sfindingsfrom the first yearof im-
plementation.
[Article copies available for a fee from The Haworth Document
Delivery Service: 1-800-HAWORTH. E-mail address: <getinfo@haworth
pressinc.com> Website: <http://www.HaworthPress.com> © 2001 by The Haworth
Press, Inc. All rights reserved.]
KEYWORDS.Bullying, sexual harassment,school violence,gender vi-
olence, school-based prevention, elementary school, victimization
The school shootings of the past two years and the recent incident in Cen
-
tral Park force us to face issues that up until now we’ve allowed to remain
somewherefarfrom theforefront of ourminds. We nolonger havetheluxury
of ignoring what has been an ugly part of our common culture and a painful
part of growing up for most American children. Bullying, sexual harassment
and gender-based violence create a range of social and academic problems
for many of our young people, beginning in elementary school and continu
-
ing through high school. They always have. The recent tragedies move us as
a nation to confront these abusive behaviors and search for ways to stop
them.
HISTORY OF THE PROJECT
SafePlace: Domestic Violence and Sexual Assault Survival Center is the
Austin/Travis County community’s primary provider of comprehensive sex
-
158 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
ual and domestic violence prevention, intervention, education and advocacy
services. Since the inception of SafePlace’s parent agencies in the mid-1970s
SafePlace has developed and provided domestic violence and sexual assault
and abuse services in response to the needs of the community. In the late
1980s, when counselors from local schools began requesting counseling for
girls in violent dating relationships, domestic violence counselors began pro
-
viding support to students at school, beginning a nationally-recognized dating
and sexual violence prevention and intervention program for boys and girls
now known as Expect Respect. After more than a decade, SafePlace has con
-
tinually provided school-based sexual harassment and dating violence preven
-
tion and intervention to Austin-area youth.
Sexual harassment and dating violence as well as the bullying behaviors
that precede such gender-based violence are rampant in schools and among
school-age youth. However, a majority of school personnel and teachers are
not properly addressingthesebehaviors with the seriousness they demand. Of-
ten teachers and school personnel are not confident in their ability to intervene
on behalf of the victim. Some school personnel discount these behaviors as in-
consequential or consider them to be flirting or teasing, leaving the victims
feeling powerless and giving the perpetrators carte blanche to abuse again.
Students, after seeing a lack of action by school personnel, do not help victims
of harassment, knowing they will not have the support of adults in charge.
Numerous studies and surveys have found that sexual harassment is a sig-
nificant problem in schools, particularly middle and high schools. A survey of
students ingrades 8-11 found that81% of the respondentsreported some expe-
rience of sexual harassment in school. Of these, 66% said they had been ha-
rassed at least once in the halls, 55% in the classroom, 43% on other school
grounds, 39% in the gym or other athletic area, 34% in the cafeteria, and 23%
in the parking lot. (AAUW, 1993) Such high rates of harassment indicate a
lack of action by school personnel. As one 14-year-old female student stated,
“I’vebeen harassedinfront ofteachers andhall monitors,maybe evenajanitor
ortwo,and certainlyotherstudents, noneofwhom tookany action. Theyprob
-
ably dismissed it as flirting, or maybe they were just ignorant or didn’t
care”(Stein, Marshall and Tropp, 1993, p. 3,5).
The AAUW survey also gauged the educational and behavioral impacts of
sexual harassment on the targets. From one-fifth to one-third of harassed girls
reported experiencing the following educational consequences: not wanting to
go to school; not wanting to talk as much in class; finding it hard to pay atten
-
tion in class; staying home from school or cutting a class; and/or making a
lower grade on a test or in a class. Behavioral consequences of sexual harass
-
ment included avoiding the person, staying away from a particular place,
changing seats in class, and withdrawing from a particular activity or sport.
Interventions and Prevention 159
State and local data also demonstrate that students experience violent and
controlling behavior at alarming rates in their dating relationships. A study of
8th and 9th grade male and female students in North Carolina indicated that
25% had been victims of non-sexual dating violence and 8% had been victims
of sexual dating violence. A 1997 Massachusetts Youth Risk Behavior Sur
-
vey, using a statewide and scientific sample of high school students, reported
that 14% of students experienced dating violence (7% males and 20% fe
-
males),11%had beenphysically hurt,and7% hadbeen hurtsexuallyby some
-
one they were dating. Of 7,000 Austin Independent School District students
who participated in educational presentations on dating and sexual violence
during the 1998-1999 school year, 11% reported past abuse by a dating part
-
ner, 3% reported having abused a dating partner, 29% had a friend that had
been abused by a dating partner, 22% had a friend that had abused a dating
partner, and 65% knew someone who had been raped or sexually assaulted
(SafePlace, Final Evaluation, 1999).
Sexual harassment, dating violence and sexual assault and abuse among
school-age youth are not phenomena that occur without warning. In fact, ag-
gressive andharassing behaviors begin muchearlieramong students in thepri-
mary grades. “The antecedents of peer-to-peer sexual harassment in schools
may be found in ‘bullying,’ behaviors children learn, practice, or experience
beginning at a very young age...Left unchecked and unchallenged, bullying
may in fact serve as fertile practice ground for sexual harassment.” (Stein,
1999, p. 50; Stein, 1993, Stein, 1995) The National Educational Association
estimates that 160,000 students miss school every day or 28 million days per
year, due to fear of attack or intimidation by a bully (Fried & Fried, 1996).
MODEL: EXPECT RESPECT ELEMENTARY SCHOOL PROJECT
In 1997, SafePlace received one of ten grants awarded by the Centers for
Disease Control and Prevention to develop, implement and evaluate an inti
-
mate partner violence prevention program in twelve Austin Independent
School District (AISD) elementary schools. The project was rooted in the be
-
liefthatsocial acceptanceof violence inpersonal relationshipsisa majorcause
and perpetuating factor in the occurrence of dating and domestic violence, and
that children exposed to a culture which accepts bullying and sexual harass
-
ment in school or domestic violence at home are at increased risk for using and
accepting violence in their own relationships. The project was designed to as
-
sist schools in addressing bullying and sexual harassment on their campuses,
which, left unchecked, condition students to accept mistreatment in their peer
160 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
relationships and lay the foundation for dating and domestic violence later in
their lives.
For the Expect Respect Elementary School Project, SafePlace conducted a
survey of 1,500 fifth graders in 12 AISD elementary schools. As expected,
many of the students had been bullied during school hours on school grounds.
Thesurveyrevealed that45% of respondentsexperienced bullying atschool or
on the bus in the preceding three months. Incidents of bullying occurred in the
classroom, in the cafeteria, on the playground, and in the bathrooms.
The Expect Respect Elementary School Project utilized a whole school ap
-
proach to improve the climate on six intervention campuses. The following
five components are described below: (1) classroom education; (2) staff train
-
ing; (3) policy and procedure development; (4) parent education; and (5) sup
-
port services
Classroom Education
All fifth grade students in the intervention schools received 12 weekly ses-
sions adapted from Bullyproof: A Teachers Guide on Teasing and Bullying for
Use with Fourth and Fifth Grade Students (Stein, 1996) facilitated by the
project’s educators and the classroom teacher or school counselor. The
Bullyproof lessons helped the students distinguish between playful teas-
ing/joking around and hurtful teasing/bullying. They enhanced students’
knowledge about bullying and sexual harassment and developed their skills
for responding to bullying and sexual harassment as targets and as bystanders
byfocusingon courage andtaking action. The lessonsincluded writing assign-
ments, role plays and class discussions and were correlated with the state’s
curriculum requirements by project staff.
Staff Training
All adults working on campus received training to raise their awareness of
bullying and sexual harassment and prepare them to respond effectively when
they witness incidents among students or have incidents reported to them.
Training included discussion of the research on bullying and sexual harass
-
ment; strategies for building a consistent response at the individual, classroom
and school-wide levels; strategies for classroom management that enhance
mutual respect among students; practice using lessons from the prevention
curricula; and methods for integrating the lessons into many areas of the cur
-
riculum, i.e., social studies,languagearts, health. Campus administrators were
urged to include all staff in training, and schools were reimbursed for the cost
of substitute teachers when required. Special training was delivered to school
Interventions and Prevention 161
bus drivers and their supervisors, as much of bullying and sexual harassment
occur on the ride to and from school. Intervention schools had a six-hour
start-uptrainingby Nan Steinand an additional threehours of trainingeach ac
-
ademic year by project staff.
Policy and Procedure Development
Campus policy and procedures were developed and adopted to facilitate a
consistent response by all staff members to incidents they witness or about
which they are told. These documents include a statement of philosophy,
working definitions of bullyingandsexual harassment, expectations for action
in response to incidents and reports, and a statement of commitment to main
-
taining confidentiality of targets, witnesses who report incidents, and students
accused of bullying or harassing. These documents were developed with input
from staff at all levels and approved for adoption by the school principal and
thecampusadvisory council (madeup of staffandparent representatives). The
principalateach campus providedtraining for all staffon the policyand proce-
dures and the documents were distributed to all staff.
Parent Education
Information about bullying and sexual harassment was offered to parents to
facilitate their support of the program and to provide them with a common lan-
guage withwhich to respond to incidentsor discussions with their childrenin a
manner consistent with the project objectives. Parent education seminarswere
offered and newsletters were sent home with children. Seminars provided in
-
formation about bullying and sexual harassment, the vocabulary being used
with students, ways to respond to and prevent sibling bullying at home, and
community resources for help with bullying, sexual harassment, gender vio
-
lence and domestic violence.
Support Services for Individual Students
Individualand group counselingwas madeavailable to studentsthrough the
school counselor. Special attention was paid to students who were routinely
bullied or who were bullying others repeatedly. The Expect Respect Project
provided school counselors with a manual of community resources and infor
-
mation about sexual harassment, bullying, and dating and domestic violence.
Counselors from SafePlace provided individual counseling to students strug
-
gling with any or all of these issues.
162 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
Partnerships
SafePlace was the lead agency of the project. As such, SafePlace was fi
-
nancially and programmatically responsible for the all aspects of the project,
from development of the design through implementation and evaluation.
SafePlace had two partners in addition to the CDC, which provided the fund
-
ing andtechnical assistance. The Universityof Texas at Austinwas responsi
-
ble for the evaluation component of the project. Staff from the Center for
Social Work Research of the School of Social Work, the School of Nursing,
and the College of Education served as technical advisors and conducted the
data analysis. The Austin Independent School District (AISD) had twelve
schools represented in the project–six intervention schools and six compari
-
son schools. In addition, Nan Stein, Senior Researcher at the Wellesley Col
-
lege Center for Research on Women served as a consultant to the project.
Stein contributed her expertise to all aspects of the project’s design and pro-
vided initial training on her Bullyproof curriculum to SafePlace and AISD
staff.
STUDY DESIGN
This study involved a pretest posttest design with four groups, two groups
(fall and spring cohorts) of intervention schools and two groups of comparison
schools. Each group included three schools. During the 1998-99 and
1999-2000schoolyears, services weredelivered and datawascollected for the
two groups of intervention schools. Data was also collected from the two
groups of comparison schools, although the comparison schools did not re
-
ceive any of the program services. All project participants were nested within
schools (Cook & Campbell, 1979).
During the two years of the project implementation, student and staff sur
-
veys conducted at the comparison and intervention schools were collected for
analysis of the effectiveness of the services at meeting the project objectives.
Staff at both intervention and comparison schools received the staff survey at
the beginning of the year and as a follow-up at the end of the school year. Fifth
grade students at the intervention and comparison schools received the student
survey at the beginning of the year, at mid school year, and at the end of the
school year. The fifth grade teachers also received the mid school year survey
along with their classes so that analysis could be done measuring the relation
-
ship between the survey responses of the teachers and the survey responses of
students in their individual classes.
Interventions and Prevention 163
Sample Selection and Size
Elementary schools were selected for this project due to the significant
amount of bullying and student-to-student sexual harassment occurring at the
elementary school level. Fifth grade students were selected for the classroom
education component for three reasons. First, as the oldest students at the ele
-
mentary schools they serve as role models for younger children. They are also
preparing to go to middle school where they may be at risk for more serious
formsof bullyingandstudent-to-student sexualharassment. Andfinally, many
fifth graders have already begun to experiment with new roles of boyfriend
andgirlfriend. Itis,therefore, atime whenthey wouldbenefit fromlessons that
address safety and respect in personal relationships.
Six pairs of schools that represent a cross-sectionoftheAustinIndependent
School District were selected. The school pairs were matched on variables in
-
cluding ethnicity, limited-English proficiency, and the socio-economic status
of students, the schools’ passing rates on the statewide academic skills test
(TAAS), total school population, and fifth grade population. Schools in each
pairwere randomlyassignedtoeitheran interventionor acomparison group.
DATA ANALYSIS METHODS–QUANTITATIVE
The study included 1243 fifth-graders from 12 elementary schools within
the Austin Independent School District (Austin, Texas). Six schools received
the Bullyproof curriculum (intervention group), three in the fall of 1998 and
three in the spring of 1999. In an attempt to have an equal demographic repre-
sentation in each of the subgroups (e.g., gender and ethnicity), each of the six
intervention schools (intervention group) was matched with one of the six
comparison schools (comparison group) that was similar on the proportion of
these variables in the school population. The intervention school was ran
-
domly selected in each matched pair of schools. The 1243 fifth grade students
completed a survey at three time points: pretest at the beginning of the school
year (Time 1), late fall (Time 2), and spring (Time 3). A total of 496 students
were eliminated from the current analysis for the following reasons: Students
were not includedintheanalysis if they did not take the survey atallthree time
points;ifthey hadcompleted the Spanishversion of thesurvey (n =20)
1
;andif
they had changed schools during the school year. Consequently, 747 students
from 12 schools were included in the current analysis
2
. Of the students in
-
cluded in the analysis, 362 (48%) were students at the intervention schools.
Duetoa smallsample size,AmericanIndian (n= 3)andAsian (n= 33)were
grouped together with White. Consequently, for analysis purposes there were
164 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
only three categories for ethnicity: African American, Hispanic, and
White/Other. Data on socioeconomic status (SES) was based on whether or
not the student and/or their sibling(s) were receiving free or reduced-price
lunches. For analysis purposes, students receiving free or reduced-price
lunches (lowSESgroup) were compared with thestudentswho did not receive
free or reduced-price lunches (non-low SES).
The intervention group consisted of 48% males and 52% females. Of those,
16% were African American, 26% Hispanic, and 58% White/American In
-
dian/Asian. Sixty-seven per cent of students were non-low SES and 33% were
low SES (see Figures 1, 2 and 3).
DATA ANALYSIS STRATEGY–QUANTITATIVE
Analyses are based on student answers to a 27-question survey that con
-
sisted of multiple-choice and yes-no questions. The two primary methods of
analysis of the student survey consisted of multilevel or hierarchical linear
models and chi-square analysis. The chi-square analyses were used to analyze
changes in student responses to the multiple-choice items on the survey. A
six-cell matrix was used to compare the number of students who changed their
answersfrom Time1to Time3 ina desirableoran undesirabledirection tostu-
dents whose answers were consistent for both survey administrations. The re-
sponses of students at intervention schools were matched to students at the
comparison schools to determine the effect of the project.
Multilevel or hierarchical linear models were fit using the SAS MIXED pro-
cedure (SAS Institute, 1996) in order to properly account for variation in bully
-
ing and sexual harassment knowledge due to school membership in determining
the project effect on bullying and sexual harassment knowledge. One of the de
-
pendent variables used in the multilevel analysis was the composite score for
bullying (see Appendix). The reliability (Cronbach Alpha) of the composite
scorewasadequate ateachof thetime points (.67,.69, & .62).The other depend
-
ent variable used in the multilevel analysis was the composite score for sexual
harassment (see Appendix). The reliability (Cronbach Alpha) of the composite
score was adequate at each of the time points (.66, .64, & .62).
Quantitative Results
Knowledge of Bullying and Sexual Harassment
Students in both the intervention and comparison groups showed slight im
-
provement in their knowledge of bullying. However, the hierarchical linear
Interventions and Prevention 165
modeling (HLM) did not indicate that the classroom education had a signifi
-
cant effect on bullying knowledge for students at the intervention schools. As
part of the hierarchical linear model analyses, four separate models were fitted
for the bullying dependent variable. Each model tests the effect of the interac
-
tion of different groupings of the variables: time, group, cohort and ethnicity
and SES.While there wereincreasesover time, noneof these models indicated
166 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
Percentage
of Students
Male
Female
44.71%
55.29%
48.34%
51.66%
Control (N = 378)
Intervention (N = 362)
60%
50%
40%
30%
20%
10%
0%
FIGURE 1. Gender of Students Included in Analysis
Percentage
of Students
Control (N = 378) Intervention (N = 362)
African American
Hispanic
White/Other
16.67%
23.54%
59.79%
15.75%
26.24%
58.01%
60%
50%
40%
30%
20%
10%
0%
FIGURE 2. Ethnicity of Students Included in Analysis
statistically significant differences between the intervention and comparison
groups resulting from exposure to the classroom education.
In contrast, the hierarchical linear modeling (HLM) indicatedthattheclass-
room education did have a significant effect on sexual harassment knowledge
for students at the intervention schools. Four separate models were also fitted
for the sexual harassment dependent variable. Descriptions of the models are
included in Table 1 as well as a table of the F-ratios, degrees of freedom, and
significance levels. The Cohort by Time interaction was determined to be sig-
nificant for sexual harassment knowledge. Therefore the change in sexual ha-
rassment knowledge over time can be partially explained by which cohort
students were in, i.e., students in the fall cohort showed greater increases in
knowledge betweenTime 1 andTime2 than thespring cohort, whohadnot yet
received theclassroom education. Therewasa statistically significanteffectof
measurement time F(2, 2204) = 113.46, p < .001. Sexual harassment aware
-
ness increased from Time 1 (M = 7.05, SE = .102) to Time 2 (M = 7.65, SE =
.098) to Time 3 (M = 7.96, SE = .097). This main effect was qualified by an in
-
tervention Group by Time interaction, F(2, 2204) = 3.59, p < .05. Figure 4 il
-
lustrates the interaction and shows that the intervention group had a larger
increase over time relative to the comparison group.
The Group by Time interaction remained statistically significant across the
four models. Gender and SES by Time also remained significant in both the
models that introduced them as variables.
Figure 5 shows the interaction by plotting the means for each cell in the in
-
teraction. The non-low SES intervention group appears to have started higher
than othergroupsand showed consistent increases from Time1to Time 3. The
Interventions and Prevention 167
Percentage
of Students
Non Low SES
Low SES
Control (N = 380)
Intervention (N = 367)
70.26%
29.74%
66.76%
33.24%
80%
60%
40%
20%
0%
FIGURE 3. SES of Students’ Families for Analyzed Group
low SES intervention group showed an increased change from Time 1 to Time
2 but then leveled off from Time 2 to Time 3. The non-low SES comparison
group shows a fairly similar increase from Time 1 to Time 2 compared to the
low SES intervention group. The low SES intervention group shows a consis
-
tent increase over time.
168 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
TABLE 1. F-Ratios and Significance Levels for Fitted Effects Models with the
Sexual Harassment Dependent Variable
Sexual Harassment
Predictors
Numerator Model 1 Model 2 Model 3 Model 4
DF DDF = 2204 DDF = 2200 DDF = 2198 DDF = 2178
Group 1 2.51 3.39 5.31* 4.26*
Time (Ntime) 2 113.46*** 107.18*** 108.51*** 85.61***
Group X Time 2 3.59* 3.45* 3.59* 3.48*
Cohort (Time) 1 - 0.00 0.06 .05
Group X Cohort 1 - 3.22 -
Time X Cohort 2 - 6.01** 6.07** 6.23**
Time X Group
X Cohort 2 - 0.36 - -
Gender 1 - - 32.31*** 32.60***
Ethnicity 2 - - 0.34 0.32
SES 1 - - 7.27** 6.83**
Gender X Group 1 - - - 0.89
Ethnicity X Group 2 - - - 0.44
SES X Group 1 - - - 1.29
Gender X Time 2 - - - 0.20
Ethnicity X Time 4 - - - 0.60
SES X Time 2 - - - 0.09
Gender X Group X Time 2 - - - 3.69*
Ethnicity X Group X Time 4 - - - 1.08
SES X Group X Time 2 - - - 3.89*
Note: DDF = Denominator Degrees of Freedom. Note * = p < .05; ** = p < .01; *** = p < .001
Model 1 consisted of the intervention group, time of measurement, and the interaction of these two predictors.
Model 2 contained the same predictors as Model 1 with the addition of the cohort (Fall versus Spring) effect.
Model 3 contained the same predictors as Model 1 with the addition of respondent’s gender (male versus fe
-
male), ethnicity (African-American, Hispanic, White/Asian/American Indian), and socioeconomic status (low SES
versus non-low SES) as main effect predictors.
Model 4 contained the same predictors as Model 3 with the addition of the interactions of gender, ethnicity, and
socioeconomic status (SES) with intervention group, time of measurement, and the intervention group by time
interaction.
Interventions and Prevention 169
Control
Intervention
Time 1
Time 2 Time 3
Sexual Harassment Score
8.5
8
7.5
7
6.5
6
FIGURE 4. Interaction Between Measurement Time and Groupfor Sexual Ha
-
rassment Score
Sexual Harassment Score
Time 1
Time 2
Time 3
Low SES Control
Low SES-Intervention
Non-low SES-Control
Non-low SES-Intervention
8.5
8
7.5
7
6.5
6
FIGURE 5. Interaction Between Time, Group (Intervention versus Compari-
son) and Social Economic Status (Free Lunch versus No-Free Lunch)
There was also an interaction of Group by Time by Gender, F(2, 2178) =
3.89, p < .05. Figure 6 shows the interaction that indicates that males in the in
-
tervention group showed a greater gain than males in the comparison group
from Time 2 to Time 3. Females in the intervention group showed a greater in
-
crease than females in the comparison group from Time 1 to Time 3.
Awareness of Bullying at School
Students were asked, “How often in the past week have you seen any bully
-
ing at school or on the bus?” Response options were “not at all,” “once or
twice,” or “almost everyday.” The difference in the change scores between the
comparison and intervention schools was statistically significant, X²(2, N =
723) = 7.00, p < .05. More students at the intervention schools reported seeing
bullying more often at Time 3 than they did at Time 1 (see Figure 7). At Time
3, 59% of all students reported that they that they had seen bullying “once or
twice” or “almost everyday” in the past week (45% and 14% respectively).
Although change scores were not significant in Year 1, it is also interesting
to note that20%of students reported at Time 3, after theintervention,that they
had bullied another student in the past week. Also, at Time 3, 37% of students
surveyed reported that they had been bullied in the past three months.
170 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
Sexual Harassment Score
8.5
8
7.5
7
6.5
6
Time 1 Time 2 Time 3
Male-Control
Male-Intervention
Female-Control
Female-Intervention
FIGURE 6. Interaction Between Time, Group (Intervention versus Compari-
son) and Gender (Females versus Males)
Behavior Related to Bullying at School
The project had the effect of increasing the students’ expectations of them-
selves, rather than that of the school staff. Students in the intervention group
indicated that following the project they were more likely to take personal ac-
tion on behalf of a target. Their responses also indicated that after the project
they were less likely to tell an adult about a name-calling incident. There was
statistically significant change for the intervention group on the question,
“What would you do if you heard students calling another student mean
names?” Fewer students at intervention schools indicated at Time 3 that they
would “tell an adult at school” about a name calling incident when compared
to thepercentage for the sameresponse item atTime1, X²(2, N=747) = 4.275,
p < 0.05 (see Figure 8).
There was also a statistically significant difference between the interven
-
tion and comparison groups on this question. At Time 3, 56% of the compari
-
son group compared to 47% of the intervention group reported that they would
tell an adult (see Figure 9). This change was statistically significant at conven
-
tional level p < 0.05, X²(2, N = 747) = 5.94.
There was also a significant difference at Time 3 for the change in reports
between the intervention and comparison groups for the response group “take
personal action,” X²(2, N = 747) = 4.12, p < 0.05. That is, a greater percentage
Interventions and Prevention 171
Percentage
of Students
Decrease
No Charge
Increase
Control (N = 366)
Intervention (N = 357)
18.30%
11.50%
51.40%
53.50%
30.30%
35.00%
55%
50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
FIGURE 7. “How Often in the Past Week Have You Seen Any Bullying at
School or on the Bus?” Change Scores for Response Options: “Not At All,”
“Once or Twice,” or “Almost Everyday”
of the intervention group compared to the comparison group indicated that
they would “tell the students to stop calling another student mean names,”
“help thestudent get away from thebully” or “hit, kickorshove the bully” (see
Figure 9).
When students were asked, “What would you do if you saw one student
beating up another student at school?” personal action again included the re
-
172 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
70%
60%
50%
40%
30%
20%
10%
0%
Time 1
Time 3
Tell School Staff Tell
Parent/Guardian
Take Personal
Action
X = (2, n = 362) = 4.274 < 0.05
2
p
FIGURE 8. “What Would You Do if You Heard Students Calling Another Stu
-
dent Mean Names?” Intervention Group: Change From Time 1 to Time 3
70%
60%
50%
40%
30%
20%
10%
0%
Tell School Staff
Tell Parent/Guardian
Take Personal Action
Comparison
Intervention
X (2, n = 747) = 5.94 < 0.05
2
p
X (2, n = 747) = 4.12 < 0.05
2
p
FIGURE 9. “What Would You Do If You Heard Students Calling Another Stu-
dent Mean Names?” Intervention Group vs. Comparison Group at Time 3
sponse options ‘tell the bully to stop,” “help the student get away from the
bully” and “hit, kick or shove the bully.” There was a significant difference for
the change in reports from Time 1 to Time 3 for the intervention groups on the
response options indicating that they would take personal action, X
2
(2,n=
362) = 16.927 p < 0.001 (see Figure 10). The change intervention students in
-
dicated in their intention to tell an adult at school or a parent about this kind of
bullying did not meet the conventional definition of statistical significance.
At Time 3, there was also a significant difference between the intervention
and comparison groups, X²(2, N = 747) = 4.80, p < 0.05. Specifically, 61% of
the intervention group compared to 49% of the comparison group indicated
they would “tell the bully to stop beating up another student.”
DATA ANALYSIS METHODS–QUALITATIVE
Focus group interviews were conducted at each of the six intervention
schools. The focus groups sessions were divided into three distinct participant
groups:girls,boys and5th grade teachers.
3
Allteacherswere invitedto partici-
pate in the focus group discussions and only one teacher chose not to partici-
pate. Students were selected at random based on previous parental permission
to participate in the evaluation of the Expect Respect Project. Students were
Interventions and Prevention 173
80%
70%
60%
50%
40%
30%
20%
10%
0%
X = (2, n = 362) = 16,927 < 0.001
2
p
Tell School Staff
Te l l
Parent/Guardian
Take Personal
Action
Time 1
Time 3
Intervention Group Change Over Time
At Time 3, there was also a significant difference between the intervention.
FIGURE10.“What Would YouDoifYou SawOneStudentBeating Up Another
Student at School?”
also given an opportunity to choose not to participate. No student exercised
that option.
The focus group sessions were audio-taped and lasted one hour (plus or mi
-
nus 15 minutes). The project evaluator facilitated all focus groups. A research
assistant was present to take notes. Teacher sessions were held after school,
with the exception of one school where arrangements were made to conduct
the focus group session during school hours. Student sessions were all held
during school hours.
Tapes were transcribed then reviewed and coded for themes. The themes
and transcripts were then reviewed again to ensure that coding was consistent
with the identified themes. A report was presented to the project evaluator by
staff at the Center for Social Work Research. The report was then compared to
the theme analysis by the project evaluator.
QUALITATIVE RESULTS
In examining the transcripts from student and teacher focus groups, two spe-
cific themes were identified related to the project’s impact on students. The
themes were: (1) students’ conceptual awareness of bullying and sexual harass-
ment; and (2) students’ behavior. The data supports the finding of greater student
awareness of bullying and sexually harassing behavior, their own and that of their
peers. Although staff and students reported that students were more likely to inter-
venewhen someonewasbeingbullied,neither reportedthat bullyinghad stopped.
Awareness of Bullying at School
Both students and teachers reported that the project increased student aware
-
ness of the behavior that constitutes bullying and sexual harassment, the poten
-
tialimpact ontargetedstudents,andstrategies forhandling it.The studentssaid:
I learned...that just because of their gender and their body parts doesn’t
meanthat youhavetomakefun ofthem, becauseit islike apartofnature.
We learned how to respect each other.
If someone is bothering you, you could be like tell the teacher, or just
walk away and ignore them, or tell an adult.
Some teachers expressed surprise at the students’ expression of empathy
and internal questioning. Teachers described the project’s effect on student
awareness with the following statements:
174 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
It helped establish clear boundaries between acceptable and unaccept
-
able behavior.
And I basically feel that the program really helped establish an aware
-
ness of the bullying concepts and sexual attitudes, and sexist attitudes
that we are dealing with in society, so I think it was very informative for
the children. I could really see them soul-searching, okay, who did I hurt
and how is that going to be perceived.
I guess maybe the surprise that might have hit me the most maybe how,
was how some of the students were kind of empathetic towards one stu
-
dent who acts out a lot in my classroom. And so, I mean, because he acts
out that he’s setting himself up to be an easy target. But yet, there was in
-
stead of ...Idon’t know attacking him or, you know, bullying him in a
sense, they, there was some empathy there.
They were made aware of each other, being able to identify other children’s
behavior. Their behavior did not necessarily improve, but Ithink the aware-
ness gave them a little more strength in not accepting bullying . . . Thebully
has not got the power that maybe they had before.
Teachers also reported that the project improved the students’ vocabulary
for describing bullying situations. Before the project, bullying reportswereof-
ten labeled as “tattling” and dismissed. Following the project, students and
teachers hada common language withwhich to define incidents,behavior,and
expectations more precisely. Many teachers reported a noticeable increase in
reports of bullying.
The students’ ability to describe bullying and sexual harassment with
specific vocabulary seemed to make them more conscious about the is
-
suesof bullyingandsexualharassment,and whatthey sayto eachother.
I think it’s been more positive in my class. It’s given us a point of refer
-
ence to deal with situations. Hasn’t necessarily gotten any better, but
we’ve dealt with it more effectively.
For me, I don’t think there were more instances, probably [sic] people
were finally telling us about it. They felt much more comfortable report
-
ing.
Likeshesaid, itwasn’t tattling anymore,it was areal problem.It’ssome
-
thing that could be dealt with. They finally came up to us and then we
would acknowledge it instead of, “go settle it on your own.”
Interventions and Prevention 175
Behavior Related to Bullying at School
While some students and teachers reported that bystander behavior did not
change, others gave specific examples of students intervening on behalf of an
-
other student who was being bullied. The students indicated in the surveys that
they were more likely to intervene on behalf of a target and, on the whole, the
focus group data supports these findings. Students said:
Before ExpectRespectI wasn’t standing up formyfriends and stuff [sic]
and they would get somebody to jump me or something. With Expect
Respect Mr. Johnson told us not to be scared and to show our feelings.
Then when they were picking on my friend, I said now leave her alone
because I got tired of them picking on her everyday . . .
Some people they taking up for people. Before Expect Respect . . . they
take up for people now . . . well sometimes.
This boy has learning problems and he has a really nice sister, a lot of
people teaseand bully him.Andnow I standup for him, nowI know who
his sister is and she’s really nice. Sometimes I feel bad because if there
wasn’t Expect Respect I probably wouldn’t stick up for him.
Many of the teachers expected that students would stop bullying each other
as a result of the Bullyproof lessons in their classrooms. The project staff did
not share this expectation due to the limited time that students would be dis-
cussing the issues during the Expect Respect sessions. Focus group data is in-
conclusive on this point. Some students and teachers reported positive change,
while others reported negative change or no change at all.In some of the focus
group sessions there were individuals who gave conflicting opinions about be
-
havior change. For example, one student said:
Some people they still fighting and fussing at each other. A whole bunch
of people, they’re not fighting and fussing at each other, but sometimes
they are. We don’t have a lot of fights in our class no more . . .
While teachers at different schools said:
It was disturbing to me at the first when I saw more bullying, it really
was.
I think awareness is probably the biggest thing, and probably the biggest
disappointment is behavior did not really change.
176 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
Teachersacross the sixschools alsoindicated thatsomestudents wereusing
mildcases ofbullyingto getattention. Manyof theteachersacknowledged this
testing of new limits as typical behavior for this age group.
I think about 2 or 3 weeks into the project they started trying out the
scenes on us, and seeing how far we would let it go if we knew it was
happening. Like they really wanted to make us aware we’re doing this,
what are you going to do about it now?
But as we mentioned, I think the other thing is that it also gave them, some...
not ideas, but it made them think, “Well, if we do this, then we’ll get this kind
of attention.” It’s almost like they are working at trying to get some of the at
-
tention now. It seems almost like deliberate choices are being made.
Teachers at one school specifically noted that the severity of bullying that
students were involved in was less than in years before the project began.
Well, actually I meant, you know, going back, for me, I’ve been thinking
about it, getting back from it though, and going back over the last couple
of years, especially the last couple years, previous to this year. I think
that [sic] our 5th graders are a lot less prone to be very hurtful.
DISCUSSION OF RESULTS
The ExpectRespect Project evaluation examinedthe project’s effect onstu-
dents’ knowledge, awareness and behavior related to bullying and sexual ha-
rassment at school. The multilevel analysis indicated that the project had an
impactonsexual harassment knowledge.Students in bothgroups,intervention
and comparison, had improvements in their scores for sexual harassment
knowledge, but students at the intervention schools demonstrated greater in
-
creases in scores for these questions than students at the comparison schools.
Knowledgeaboutbullying, however, wasnot significantly affectedby the pro
-
ject according to the multilevel analysis. This can be partially explained by the
fact that students had relatively high initial scores on the bullying knowledge
questions. It is also worth mentioning that the response pattern for students at
the comparison schools suggest that repeated exposure to the project survey
may have had some teaching effect for the comparison school students.
Both the chi-square and focus group data analyses indicate that the students
developed greater awareness of concepts and recognition of bullying in their
environment. In particular, teachers said that the students’ increased vocabu
-
lary made it much easier to deal with bullying incidents when they occurred.
Interventions and Prevention 177
When approaching the teachers about bullying situations the students were
better able to articulate the specific problems.
The chi-square analyses indicate that students at the intervention schools
were significantly more likely to identify bullying more often and in more
places at their school after the intervention. The students at the intervention
schools were also more likely to take personal action on behalf of a target of
bullying when compared to the students at the comparison schools. Students
and teachers indicated in focus groups that they saw some increased interven
-
tion by fifth grade students in bullying situations.
Important project effects are indicated in three areas of student growth and
development:sexual harassmentknowledge;awareness ofbullying andits’ ef
-
fects; and willingness to intervene on behalf of someone being bullied. One
unintended effect of the project as reported by some teachers was an increase
in mild bullying by some students. However, teachers indicated much of this
increase seemed to be either attention-seeking behavior or normal testing of
new limits typically demonstrated by students at this age.
Successes
The Expect Respect Elementary School Project successfully increased stu-
dent’s knowledge of sexual harassment, their awareness of bullying on cam-
pus, and the likelihood of them intervening on behalf of a target. The project
also increased public awareness through local workshops, media exposure,
and through presentations at state and national conferences. As the nation has
recognized the relationship betweenbullyingandschool violence, requests for
help have also increased. In Austin, ten schools that did not participate in the
research project requested and received student and staff training from project
staff. The school intends to support K-12 prevention education district wide,
and SafePlace continues working with the district to develop and implement
student-to-student sexual harassment policy and training.
Challenges
A primary challenge to the success of the project was it’s applicability to
Latino students and their families. Austin’s school population is 41% Latino.
There are many recent immigrants among this group, and many monolingual
Spanish-speaking students. There is no term in Spanish that is equivalent to
bully or bullying in English. Neither the term nor the concept translates accu
-
rately. Difficulty also lies in cultural differences. Gender role expectations
vary from culture to culture and with them so do definitions of healthy rela
-
178 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
tionships. This project did not have the resources to adequately address these
differences.
The other substantial challenge was getting sufficient time allotted for
teachers to be adequately trained on the use of the curriculum and on effective
responses to incidents. The emphasis in Texas on standardized test scores has
intensified the pressure on teachers and administrators to use all available time
for test preparation. As a result, getting a commitment of time for training was
extremely difficult.
RECOMMENDATIONS
Addressing bullying and sexual harassment on school campuses requires
fullsupportof thecampus staff.Asmentioned earlier,two factorsthatare criti
-
cal to changing the school climate from one that allows mistreatment to one
thatpromotesrespect amongstudentsare: (1)theadults intervene immediately
and in a consistent manner when they witness bullying or sexual harassment,
and (2) students are prepared and motivated to take personal action on behalf
of a target. The key element in creating this change is the support and invest-
ment of the school’s principal. S/he establishes expectations for staff behavior
and sets priorities for their time and the focus of their attention. A school-wide
commitment to the project’s goals, as established by the principal, also deter-
mines the continuation of activities after the project ends. For a commu-
nity-based agency to work effectively in schools, time and resources must be
allocated to gaining the commitment and full investment of school principals.
The project director must be knowledgeable of, and sensitive to school culture
and the logistical and political pressures on the administrators and teachers. In
the case of the Expect Respect Elementary School Project, a strong relation
-
ship between SafePlaceand the AISD laid the foundation forasuccessful part
-
nership.
NOTES
1. The difficulty in administration of Spanish language surveys made the validity of
their results highly questionable.
2. There were significant differences between the 496 excluded and the 747 in
-
cluded students, interms of student’sgender, X²(1,N= 1186) =5.27, p<.05; ethnicity,
Interventions and Prevention 179
X²(2, N = 1186) = 18.4, p < .001; and SES, X²(2, N = 1233) = 17.63, p < .001. There
were fewer males, more females, fewer African Americans and Hispanics and more
Whites,andmore non-low SES studentsintheincludedgroupthan the excluded sample.
3. Originally the sessions were divided into only two groups, students and 5th grade
teachers. The decision to separate the girl and boys was made after the first student fo
-
cus group session.
180 BULLYING BEHAVIOR: CURRENT ISSUES, RESEARCH, AND INTERVENTIONS
APPENDIX
Questions on Bullying and Sexual Harassment Knowledge
Students were asked to answer the following two questions to assess their bullying and
sexual harassment knowledge. Students received composite scores for each question. The
composite scores werethen used in the multilevel analyses to determinethe project effecton
each knowledge base.
Which of these would you call bullying?
Check as many as you call bullying
q Name calling
q Pushing someone
q Telling someone to leave you alone
q Crowding or cornering someone
q Making fun of a person’s body
q Telling on someone who did something wrong
q Hitting someone
q Telling mean jokes about someone
q Taking things from someone without asking
q Not sharing your lunch or snack
q Shooting the finger at someone
q Not telling someone about your birthday party
q None of the above
q Other (please write it in)
Which of these would you call sexual harassment?
Check as many as you call sexual harassment
q Touching someone in ways that are not OK with him or her
q Making fun of someone’s private body parts
q Having a dream about someone you like
q Passing a note that says sexual things about someone’s body
q Continuing to tell dirty jokes around someone after he or she has asked you to stop
q Letting someone know that you like her or him
q Asking for someone’s phone number
q Asking the counselor for information about sex
q Pressuring someone for sex
q None of the above
q Other (please write it in)
